3D Modeling Tips and Tricks

I've been building a ton of loudspeakers and waveguides in 3D, and thought it might be worthwhile to post some of the shortcuts I've figured out.

It took me a while to figure out how to quickly make some of these shapes, so maybe this will save people some time.

First, here is how I print:

1) I use Autodesk 123D Design to model the speakers ($0)
2) I use Repetier and Slic3r to 'cut up' the model into something that can be printed ($0)
3) I use a Printrbot Simple Metal to print the designs ($599)

It is true that you can have things printed by Shapeways. I prefer printing it myself because you can get results in about twelve hours instead of a few days. And filament is so cheap, I can print a nice waveguide for about five bucks.
 
Object one : an elliptical waveguide

3D-howto-01.jpg

The first step in making an elliptical waveguide is to draw an ellipse. In this case, I am using the golden ratio. The ellipse has a width of 12.7mm and a height of 20.5mm. When we're finished this will give us a 1" wide entrance to the waveguide.

3D-howto-02.jpg

Step two is to make the waveguide walls. I'm making a progressive transition waveguide. According to JBL, "Progressive Transition waveguides are unique because a single mathematically-continuous surface defines the waveguide from transducer- throat to waveguide-mouth. The distinctive feature is the lack of a traditional diffraction slot. Instead the sidewalls transition smoothly from the driver throat through to the square or rectangular mounting flange."

Translated into English, what we have is a curve that slooooooooowly transitions from the angle at the entrance (90 degrees) to the angle at the exit (180 degrees.)

oh3vU.jpg

here's the real deal, for comparison

Note that we're not limited to an entrance of 90 and an exit of 180. We can use any angle we want. For instance, if you were using a compression driver instead of a dome, you might use a narrower entrance angle. IE, if your compression driver had an exit of twelve degrees, you could match the exit angle of the compression driver to the entrance angle of the waveguide to reduce diffraction.

The angle that you'll get in your waveguide is determined by the tangent on that circle that I cut out there. Again, I picked 45 degrees, because it works for me. You could pick any angle you want.

3D-howto-03.jpg

Once we have that curve, we need to 'scale it up' to whatever size we want in the real world. In my case, I wanted a baffle width of 8". That means that I have to scale up the width of that curve to 3.5". (3.5" for the waveguide walls, times two, plus the width of the entrance, gives us 8" total.)

Once the curve is scaled up to size, we move the curve so it's adjacent to our ellipse that we made in step one.

3D-howto-04.jpg

Then we "sweep" the waveguide wall around the ellipse. Basically the "sweep" tool lets us take one shape (the wall) and "sweep" it around another shape (the ellipse, which will be our throat.)

An externally hosted image should be here but it was not working when we last tested it.

3D-howto-05.jpg

The last step in making the waveguide is to 'squash' one dimension. Basically we take the whole waveguide and squash it down to 61.7% of it's original dimension. What this does is it gives us a nice elliptical mouth, with a golden ratio of height to width. This isn't just for cosmetics; it improves the frequency response by reducing the on-axis 'hole' that you get with a round waveguide. (A round waveguide has a dip on axis, due to everything being equidistant from the throat; by making the waveguide elliptical we fix that.)

A couple of other advantages of an elliptical waveguide is that it allows for tighter spacing between the midrange and tweeter, which improves the vertical polars. And studies from Harman indicate that people prefer wide horizontal directivity and narrow vertical directivity. (I should really provide a citation for that, but I can't find it right now.)
 
Last edited:
3D-howto-06.jpg

elliptical cutouts are difficult to cut, so here's how to make a baffle. First, we make a rectangular solid. Then we 'snap' the waveguide to the rectangular solid. Using the 'snap' tool is fast way to line up a couple of shapes; it's way more accurate and easy than trying to do it with a mouse. Once they're 'snapped' together, we use the 'ungroup' tool to unglue them from each other.

3D-howto-07.jpg

Then we merge the two shapes on top of each other. I've made an extra copy of both shapes because I'll need the copy in a minute...

3D-howto-08.jpg

In this pic, I've 'subtracted' the waveguide shape from the rectangular solid. This gives me an inverse copy of the waveguide, basically a solid that's comprised of the empty space inside of the waveguide. In the pic, you can see how perfectly smooooooth the transition is from the throat to the mouth.

3D-howto-09.jpg

3D-howto-10.jpg

Then I use that inverted shape from the last step to 'slice' a shape out of the rectangular solid. I wind up with a baffled waveguide that's 200mm x 129mm x 36mm, or 7.87" x 5.07" x 1.4"

These dimensions are just about perfect for a nice two way with a 7" woofer. The waveguide will control directivity down to 1687hz, where you can 'hand off' to the woofer. The depth of the waveguide will give us a bit of horn loading down to around 2700hz. (Check with hornresp to get the exact figure.) If you wanted to load the tweeter lower, just use a deeper waveguide.
 
Object Two : A Lecleach Horn

3D-howto-11.jpg


3D-howto-12.jpg


3D-howto-13.jpg


3D-howto-14.jpg


Here's a LeCleach horn, built using the same methods as the waveguide from the original post in the thread. To get the curve, I modeled the loudspeaker in Hornresp, then used the 'export' function to generate the width of each step in the LeCleach curve.

For me, the easiest way to plot the curve in 123D was to use the 'Polyline' tool. The Polyline tool lets you input an angle for your line. And Hornresp lists that angle when you export the horn curve. So you just take the number from Hornresp and plug it into 123D.
 
I've really become a fan of diffraction horns lately. I do a lot of Synergy Horns, and a diffraction slot really simplifies a Synergy Horn. Here's why:

The way that a horn or a waveguide works is that it 'concentrates' the output of the driver into a narrower beam. For instance, a dome tweeter on a baffle might be able to generate 90dB at 2khz on a flat baffle, but if you put it on a 90 degree waveguide, the output will go up to as high as 100, even 102dB

So... What does this have to do with Synergy Horns?

I'd argue that the hardest part about building a Synergy Horn is getting the stupid tweeter to reach low enough to 'hand off' to the midranges. I can't even count the number of Synergy Horns that I've had to toss into the trash because the tweeter 'ran out of steam' and couldn't dig deep enough to 'meet up' with the midrange.

This isn't a problem for conventional two ways, because in a conventional two way it's pretty trivial to raise the xover point on the midrange to accomodate the tweeter. But in a Synergy Horn, it's a HUGE issue, because both the mids and the tweeter are horn-loaded. Due to that, the maximum we can run the mids can be as little as 1200-1500Hz. It places a huge demand on the tweeter, which is probably one of the reason that the Danley Genesis horns use a complex device to 'sum' multiple tweeters, and many of the Danley Synergy Horns use coaxial compression drivers that cost about $1000 per pair.

So...

A diffraction slot solves a lot of these issues. Basically lets us have our cake and eat it too. It allows us to have a waveguide with a specific pattern, like 90 x 60, along with the 'gain' of a horn. (Because the diffraction slot adds about 1-2" to the depth of the horn, it loads everything on the horn to a lower frequency, the tweeter in particular.) Best of all, that diffraction slot gives us a mighty convenient spot to mount some midranges to the horn.

I'm really becoming a fan.

Anyways, enough of my proselytizing, here's how to make one.

waveguide-614-1.jpg

18Sound XT1086 is the finest diffraction horn I've measured. They put a lot of effort into reducing diffraction at the edge of the diffraction slot; you wouldn't even know it was there if you didn't look closely. I've made some attempts to copy the JBL M2 diffraction horn, and the 18Sound performs better than any JBL M2 clones I've made. It's a simple elegant design.

5dlNomo.png


BxDniHR.png

To make it, follow the same recipe in this thread that's for asymmetrical waveguides. The 'catch' is that we're making TWO asymmetrical waveguides, and we're bonding the two together.

The first piece is the diffraction slot. That slot is an asymmetrical waveguide with 40 degrees of vertical coverage and 13 degrees of horizontal coverage. And then the SECOND waveguide has 90 degrees of horizontal coverage, and 55 degrees of vertical.

The net effect is that you wind up with a waveguide that has an average coverage of 72.5 degrees, 90 degrees of horizontal coverage, 55 degrees of vertical coverage. Just like any conventional waveguide. The addition of the diffraction slot makes the overall device 1.5" deeper.

This additional depth makes a MASSIVE DIFFERENCE in loading the tweeter. Here's an example:

The 18Sound XT1086 is 5" deep. That will load the tweeter down to approximately 675Hz. (speed of sound / 5" / 4)

Now if you were to remove the diffraction slot, the depth of the waveguide would drop to 3.5" deep. That would load the tweeter to just 964hz.

That probably doesn't sound like a big difference, we're talking about half an octave. But it's coming RIGHT at a point where we need all the gain that we can get when we're building a Synergy Horn. Output from the midranges is rarely a problem, but output from the tweeter is a huge issue, and that additional depth can give us as much as 10dB more output at 1000Hz. This REALLY simplifies the crossover from midrange to tweeter in a Synergy Horn.

Ju3nceb.png

In this side view, I've sliced the waveguide in half, illustrating the vertical coverage of the waveguide. Note the transition is fairly smooth, to reduce diffraction. I could make it a lot smoother with a 'smoothing' function, but that would obscure how I built the waveguide, which is the point of this thread.

DF7tacX.png

In this vertical cutaway, the diffraction slot is more apparent.

zmuhnpC.png


qClabIF.png


The finished diffraction horn

mf1001-rcf-eminence-18sound.jpg

Another pic of the 18Sound for comparison's sake
 
In a discussion on Facebook, someone asked me to share my T/S params that I use for modeling high frequency waveguides in Hornresp.

To be honest, though I've designed a LOT of waveguides, I generally don't model the waveguides in Hornresp. The reason is because waveguides are fairly predictable in my experience. The process of making a waveguide is simple:

1) determine the vertical beamwidth

2) determine the horizontal beamwidth

3) determine whether you want a circular or rectangular mouth

4) pick a profile (conical, oblate spheroidal, tractrix, spherical, etc)

5) Match the exit angle of the compression driver to the entrance angle of the waveguide

6) Add an appropriate roundover at the mouth

And you're done! I don't see any real need to model it in hornresp because what would that accomplish? I know that the depth of the waveguide needs to be sufficient to load the driver down to about an octave below the xover frequency, but I can do that math in my head. (speed of sound / depth of waveguide / 4)

Occasionally I'll fire up axidriver to run a sim; axidriver can show you the polar response and the frequency response of a horn. But generally I only do that if I'm including some type of phase plug in the waveguide design. And even in THAT case, I'll frequently just design, build and print the phase plug without simulating it. (You can design a phase plug in about 30 minutes.)
 
diffraction slot throat distortion

I've really become a fan of diffraction horns lately. I do a lot of Synergy Horns, and a diffraction slot really simplifies a Synergy Horn. Here's why:

The way that a horn or a waveguide works is that it 'concentrates' the output of the driver into a narrower beam. For instance, a dome tweeter on a baffle might be able to generate 90dB at 2khz on a flat baffle, but if you put it on a 90 degree waveguide, the output will go up to as high as 100, even 102dB

So... What does this have to do with Synergy Horns?

I'd argue that the hardest part about building a Synergy Horn is getting the stupid tweeter to reach low enough to 'hand off' to the midranges. I can't even count the number of Synergy Horns that I've had to toss into the trash because the tweeter 'ran out of steam' and couldn't dig deep enough to 'meet up' with the midrange.

This isn't a problem for conventional two ways, because in a conventional two way it's pretty trivial to raise the xover point on the midrange to accomodate the tweeter. But in a Synergy Horn, it's a HUGE issue, because both the mids and the tweeter are horn-loaded. Due to that, the maximum we can run the mids can be as little as 1200-1500Hz. It places a huge demand on the tweeter, which is probably one of the reason that the Danley Genesis horns use a complex device to 'sum' multiple tweeters, and many of the Danley Synergy Horns use coaxial compression drivers that cost about $1000 per pair.

So...

A diffraction slot solves a lot of these issues. Basically lets us have our cake and eat it too. It allows us to have a waveguide with a specific pattern, like 90 x 60, along with the 'gain' of a horn. (Because the diffraction slot adds about 1-2" to the depth of the horn, it loads everything on the horn to a lower frequency, the tweeter in particular.) Best of all, that diffraction slot gives us a mighty convenient spot to mount some midranges to the horn.

I'm really becoming a fan.

Anyways, enough of my proselytizing, here's how to make one.

waveguide-614-1.jpg

18Sound XT1086 is the finest diffraction horn I've measured. They put a lot of effort into reducing diffraction at the edge of the diffraction slot; you wouldn't even know it was there if you didn't look closely. I've made some attempts to copy the JBL M2 diffraction horn, and the 18Sound performs better than any JBL M2 clones I've made. It's a simple elegant design.

5dlNomo.png


BxDniHR.png

To make it, follow the same recipe in this thread that's for asymmetrical waveguides. The 'catch' is that we're making TWO asymmetrical waveguides, and we're bonding the two together.

The first piece is the diffraction slot. That slot is an asymmetrical waveguide with 40 degrees of vertical coverage and 13 degrees of horizontal coverage. And then the SECOND waveguide has 90 degrees of horizontal coverage, and 55 degrees of vertical.

The net effect is that you wind up with a waveguide that has an average coverage of 72.5 degrees, 90 degrees of horizontal coverage, 55 degrees of vertical coverage. Just like any conventional waveguide. The addition of the diffraction slot makes the overall device 1.5" deeper.

This additional depth makes a MASSIVE DIFFERENCE in loading the tweeter. Here's an example:

The 18Sound XT1086 is 5" deep. That will load the tweeter down to approximately 675Hz. (speed of sound / 5" / 4)

Now if you were to remove the diffraction slot, the depth of the waveguide would drop to 3.5" deep. That would load the tweeter to just 964hz.

That probably doesn't sound like a big difference, we're talking about half an octave. But it's coming RIGHT at a point where we need all the gain that we can get when we're building a Synergy Horn. Output from the midranges is rarely a problem, but output from the tweeter is a huge issue, and that additional depth can give us as much as 10dB more output at 1000Hz. This REALLY simplifies the crossover from midrange to tweeter in a Synergy Horn.

Ju3nceb.png

In this side view, I've sliced the waveguide in half, illustrating the vertical coverage of the waveguide. Note the transition is fairly smooth, to reduce diffraction. I could make it a lot smoother with a 'smoothing' function, but that would obscure how I built the waveguide, which is the point of this thread.

DF7tacX.png

In this vertical cutaway, the diffraction slot is more apparent.

zmuhnpC.png


qClabIF.png


The finished diffraction horn

mf1001-rcf-eminence-18sound.jpg

Another pic of the 18Sound for comparison's sake

using a difraction slot with a lower flare rate will hugely increase throat distortion due to adiabatic air compression. one of the main benefits of the synergy horn is that using a conical horn means a very high flare rate where the compression driver is mounted - allowing the compression driver to operate at full output without the usual distortion levels found with other horns.
 
using a difraction slot with a lower flare rate will hugely increase throat distortion due to adiabatic air compression. one of the main benefits of the synergy horn is that using a conical horn means a very high flare rate where the compression driver is mounted - allowing the compression driver to operate at full output without the usual distortion levels found with other horns.

True, but I still like diffraction slots.

They're useful if you're trying to 'have your cake and eat it too', basically allowing wide beamwidth while still loading the tweeter.

My waveguides are for home use and my midranges are two inches in diameter; throat distortion isn't a huge concern.
 
Thanks very much for sharing this. While I have no plans to print waveguides right now, I'm sure I will at some point and I appreciate your generosity in sharing some shortcuts.

One question. You mention that an advantage of an elliptical waveguide is that it reduces ctc spacing and also provides wider horizontal dispersion and narrower vertical dispersion. That seems intuitive looking at them, but doesn't a horizontally oriented elliptical horn do the opposite? More narrow horizontal and wider vertical? Perhaps I am mistaken.
 
Thanks very much for sharing this. While I have no plans to print waveguides right now, I'm sure I will at some point and I appreciate your generosity in sharing some shortcuts.

One question. You mention that an advantage of an elliptical waveguide is that it reduces ctc spacing and also provides wider horizontal dispersion and narrower vertical dispersion. That seems intuitive looking at them, but doesn't a horizontally oriented elliptical horn do the opposite? More narrow horizontal and wider vertical? Perhaps I am mistaken.

That's correct.

A horn raises the output of a speaker by focusing it into a narrower beam. For instance, if you take a 1" dome tweeter and put it on a waveguide, the SPL level below 13,500Hz will become louder on axis.

Vifa-XT19-waveguide-FR.gif


For instance, here's a 3/4" tweeter on a waveguide. The seven decibel rise below 13.5khz is because the sound is radiating into a narrower beam. Measurement courtesy of zaphaudio.com


JBL_LSR2300_01-5PthRv5oaXY38IVUfQzf3CkoGEIt0Guq.jpg


One way to "have your cake and eat it too" is to use a wide horizontal beamwidth and a narrow vertical beamwidth.

So you get horizontal polar response that's fairly similar to a conventional dome, while also getting an increase in on-axis SPL, along with less sound 'splashed' against the floor and the side walls.
 
I'm sorry, I said it backwards. A horizontally oriented horn should give wide vertical dispersion and more narrow horizontal. I think of the horn exit as a analagous to driver size. Large horn mouth narrows directivity through cancelation occurring off axis, narrow horn exit gives less beaming the same way a smaller tweeter beams less. Did I dream this phenomenon?
 
The walls of a waveguide determine the beamwidth, for the most part.

Wide walls yield wide beamwidth, narrow walls yield narrow beamwdith.

All drivers beam, but a small tweeter beams higher in frequency. For instance, a one inch tweeter begins to beam about 13,500Hz, while a 3/4" tweeter begins to beam about 18,000hz.

1_6884e144-da60-4157-a360-bbef803495ab_1799x1349.jpg


Even then, the beamwidth is still determined by the "waveguide", which in the case of a conventional loudspeaker is the baffle of the loudspeaker itself.

Note: a waveguide cannot control the beamwidth when the wavelengths are smaller than the diameter of the waveguide. For instance, 15khz is 0.9" long. This means that 15khz doesn't "see" the walls of the waveguide, the sound just sails right through. But 1500Hz is 9" long, so it "sees" the walls of the waveguide.
 
Last edited:
Four years ago, when I wrote this thread, it took me about 2-16 hours to make a waveguide, depending on how complex it was.

Forum member "mabat" basically automated all of that, and now it is possible to make a waveguide in literally minutes. You can make a waveguide in your lunch hour.

There's a few pre-reqs to make a waveguide with mabat's software:

1) you need his software. Free at Acoustic Horn Design – The Easy Way (Ath4)
2) To clean up the "STL" file that his software generates, you need something to do that. I am using meshmixer, free at www.meshmixer.com)
3) In order to add a throat and a baffle to the 3D model, you'll need some type of 3D design program. I'm using 123D from Autodesk. For some reason Autodesk killed it, but I've heard Fusion is good (and also free.) 123D Apps & Products | Autodesk
4) You'll need a 3D printer. I am using this one : Monoprice Maker Select Plus 3D Printer - Monoprice.com
 
  • Like
Reactions: 1 users
efEMHXV.png


We create a config file for mabat's software "ath4", and then we feed it with the text file that defines our waveguide.

My text file is below. It defines a waveguide measuring 160mm x 304mm x 87mm. It has a coverage angle of 110 degrees by 68 degrees. (golden ratio.)

Note that the beauty of ath4 is that you can juggle these parameters to your hearts content. For instance, sometimes I'll find that adjust a single parameter by as little as 10% will make a huge difference in the frequency response and polar response. The great thing about ath4 is that you can figure these things out in minutes instead of days.


; -------------------------------------------------------
; Horn Geometry Definition
; -------------------------------------------------------

ThroatDiameter = 12.7 ; [mm]
ThroatAngle = 0 ; [deg]
Coverage_Horizontal = 110 ; [deg]
Coverage_Vertical = 68 ; [deg]
Depth = 87 ; [mm]
Depth.ConicSectionPart = 0.5 ; 0.5
Shape = raw2rect ; raw | raw2rect
Shape.FixedPart = 0.2 ; 0.2
Shape.CornerRadius = 50.8 ; [mm]
SEExp = 2.3 ; superellipse exponent

Termination = baffle ; baffle | free-air
Termination.Angle = 82.5 ; (for baffle; default = 90 deg)

; -------------------------------------------------------
; Mesh Setting
; -------------------------------------------------------

Mesh.AngularSegments = 56
Mesh.DepthSegments = 28
Mesh.LipSegments = 6
Mesh.CornerSegments = 5

Mesh.ThroatResolution = 5.0 ; [mm]
Mesh.WallResolution = 25.0 ; [mm]
Mesh.InterfaceResolution = 10.0 ; [mm]
Mesh.IBInterfaceRadiusRatio = 2.0
Mesh.InterfaceStrip = 10.0 ; [mm]

; -------------------------------------------------------
; ABEC Project Setting
; -------------------------------------------------------

ABEC.RadiationConditions = IB ; IB | box
ABEC.f1 = 500 ; [Hz]
ABEC.f2 = 16000 ; [Hz]
ABEC.NumFrequencies = 61
ABEC.MeshFrequency = 8000 ; [Hz]

ABEC.Polars.Distance = 3 ; [m]
ABEC.Polars.Step = 7.5 ; [deg]
ABEC.Polars.Points = 7

ABEC.Polars.Horizontal = yes
ABEC.Polars.Vertical = yes
ABEC.Polars.Diagonal = no
ABEC.Polars.DiagonalInclination = 0.0 ; 0.0 -> automatic angle for max radius

; -------------------------------------------------------
; Output
; -------------------------------------------------------

;Out.DestDir = "D:\Horns" ; current directory by default

Out.Write_STL = yes
Out.Write_MSH = no
Out.Write_ABECProject = yes

 
This post is the most important part of this entire process. If even ONE PERCENT of this post makes no sense, please let me know. If this part of the process is screwed up, all the rest will fail.

Here's why this step is so important:

Mabat's software doesn't produce a solid. It produces a surface of a waveguide. ABEC is perfectly happy to simulate that, but if you want to actually produce something in the real world, you have to add some "depth" to that surface. The surface that ath4 produces is infinitely thin, there's no depth whatsoever.

The really gnarly thing about that, is that there's about a million ways to "add depth" to that 3D surface. And how you go about it has huge implications for your loudspeaker. In particular, if you use the methods that are listed online, on other websites, you're not going to produce a great waveguide. The reason for this is that the "usual methods" tend to change the geometry. This isn't a huge problem if you're doing 3D prints of Batman or Mickey Mouse, but when you're making a waveguide, the throat must be absolutely flawless. When making an acoustic waveguide, even an error of one millimeter makes a measurable difference.

I tried a dozen different methods to do this, and this seemed to be the method that preserved the throat as flawlessly as possible.

If someone can think of a better way, I'd love to hear it, in particular it seems like it would be ideal if you reversed the normals and basically turned the shape "inside-out." But I couldn't get that to work.

Here's what DOES work:


O8XaVM0.png


WJMKxlm.png


The first step is that we fire up "meshmixer" and import the "STL" file that's been produced by "ath4"

3kfJU6X.png

In order to manipulate our STL file, we type "ctrl-a" to select all of it...

2FkP7eP.png

Under the "edit" menu in meshmixer, select "extrude"

ZJnrggv.png

Here's where things get messy. There's about a million ways to extrude this shape, but 95% of them produce a solid that's utterly unusable.

What I found that works is to first select "y axis" for the direction. What this means is that it's going to push our shape UP and turn it into a solid. (more on that later.) There's some other options there, such as "normal" and "constant." From what I can see, these make a pretty shape that might be fine for something decorative, like a Batman action figure, but we're making waveguides. The throat must be flawless. All the other options messed up the throat.

One thing that's super confusing here is that the axis that you must "push" will depend on the orientation of your STL file. For ath4 and meshmixer, I had to "push" the Y axis. If your stl file doesn't "push" correctly, you might mess around with "pushing" the X or the Z axis, and also experiment with using positive or negative numbers.

Apologies in advance if this is hideously confusing. In a nutshell, if you "push" the Y axis, and you use a positive offset of 12.7mm (one half inch) you're going to produce a nice looking solid.

Also, that offset is important too. I am personally using an offset of 12.7mm because I'm a fan of 3D printed waveguides that have thick walls. Basically my mindset is that a rigid waveguide is a great waveguide. And since we're 3D printing these things, I don't see any good reason to make the walls thin. I can see how thin walled waveguides are compelling if you're making a waveguide out of solid aluminum or plastic, like Eminence or 18Sound, but for a 3D printed waveguide, I'm a big fan of thick walls.

JvvbYRn.png

Earlier I mentioned that we're making a solid... But it's not really a solid yet. It LOOKS solid on our computer monitor, but it's not. We have to CONVERT it to a solid. We do that by selecting "all" with a ctrl-A, and then selecting "convert to solid."

7gwDKTK.png

Once that we've completed all the steps above, we'll have an object with thickness that can be expoprted as an STL. ("file" and then "export" in Meshmixer.) Basically we've taken a waveguide SURFACE and converted it into a SOLID.

At this point, we could print the thing, but we don't want to do that... It needs a baffle and a proper throat.

Again, if anyone has questions about this, please ask me. I went down a dozen dead ends before I came up with this.

Another thing, you might find that it's beneficial to reduce the number of polygons. I don't need to, for my model, but if you do, it can be done with meshmixer or meshlab.
 
Last edited:
The next step is to import the STL file into your 3D program of choice. We're doing this so that we can add a baffle and a mounting plate for the tweeter. Here's how I do it in 123D.

G1pepUH.png


WZ9gMWn.png


Step 1, we import the STL file

jli8SOc.png


Fw1RN19.png


Step 2: Convert the mesh to a solid. This allows us to manipulate it in our 3D program. In 123D, we do that by typing "m."

25WCUei.png


Here's a cutaway of our solid. Note that it's been extruded by 12.7mm. This goes back to this step : 3D Modeling Tips and Tricks

That extrusion screws up our throat, but we can fix that.